tensorflow - tensorflow.python.framework.errors_impl 在恢复模型后失败

Question

尽管查看了许多其他 StackOverflow 页面，但很难使用 TensorFlow 1.15 恢复已保存的模型

基本上，它在调用pred = model.predict(data)后失败

def restoreModel(data, input_dims, network_settings,file_path_final):
    tf.compat.v1.reset_default_graph()
    try:    
        config = tf.compat.v1.ConfigProto()
        config.gpu_options.allow_growth = True
        sess =  tf.compat.v1.Session(config=config)
        init = tf.compat.v1.global_variables_initializer()
        with tf.Session() as sess:
            sess.run(init)
            saver = tf.compat.v1.train.import_meta_graph(file_path_final+'Losv2_311-800.meta')
            saver.restore(sess, tf.train.latest_checkpoint(file_path_final))
            model = Model_DeepHit(sess, "DeepHit", input_dims, network_settings)
            pred = model.predict(data)

class Model_DeepHit:
    def __init__(self, sess, name, input_dims, network_settings):
        self.sess               = sess
        self.name               = name

        # INPUT DIMENSIONS
        self.x_dim              = input_dims['x_dim']

        self.num_Event          = input_dims['num_Event']
        self.num_Category       = input_dims['num_Category']

        # NETWORK HYPER-PARMETERS
        self.h_dim_shared       = network_settings['h_dim_shared']
        self.h_dim_CS           = network_settings['h_dim_CS']
        self.num_layers_shared  = network_settings['num_layers_shared']
        self.num_layers_CS      = network_settings['num_layers_CS']

        self.active_fn          = network_settings['active_fn']
        self.initial_W          = network_settings['initial_W']
        self.reg_W              = tf.contrib.layers.l2_regularizer(scale=1.0)
        self.reg_W_out          = tf.contrib.layers.l1_regularizer(scale=1.0)
        self._build_net()

    def predict(self, x_test, keep_prob=1.0):
        return self.sess.run(self.out, feed_dict={self.x: x_test, self.mb_size: np.shape(x_test)[0], self.keep_prob: keep_prob})

输出
Traceback（最近一次调用最后）：
文件“C:\tools\Python\Python37\lib\site-packages\tensorflow_core\python\client\session.py”，第 1365 行，在 _do_call
return fn(*args)
文件中C:\tools\Python\Python37\lib\site-packages\tensorflow_core\python\client\session.py”，第 1350 行，在 _run_fn 目标列表，run_metadata)
文件“C:\tools\Python\Python37\lib\site- packages\tensorflow_core\python\client\session.py"，第 1443 行，在 _call_tf_sessionrun run_metadata)
tensorflow.python.framework.errors_impl.FailedPreconditionError: Attempting to use uninitialized value DeepHit/fully_connected_2/biases_1
[[{{node DeepHit/fully_connected_2/biases_1 /读}}]]


注意：DeepHit 类还有其他方法，但似乎缺少一些变量来提供类，但即使查看相关的 TensorFlow 1.15 版本，我也无法弄清楚

x_test: [[ 0.63193141 -0.19093631 1.32147613 1.5206112 -0.32124657 1.09037371 -0.56493425 0.94727136 0.27852626 -1.39992279 -0.77934678 -0.01374535 -0.19709776 -0.37450909 -0.43395308 -0.17253745 0.21124831 -0.85609694 -0.17252606 -0.37291207 -1.08927999 -0.18948054 -0.15836433 -0.28718442 -0.14664455]]

输入：Tensor("DeepHit/inputs_1:0", shape=(?, 133), dtype=float32)
num_layers: 2
h_dim: 300
h_fn: <function elu at 0x00000153294DE8C8>
o_dim: 300
o_fn: <function elu at 0x00000153294DE8C8
> : <function variance_scaling_initializer.._initializer at 0x0000015340DFFAE8>
keep_prob: Tensor("DeepHit/keep_probability_1:0", shape=(), dtype=float32)
w_reg: <function l2_regularizer..l2 at 0x000001534204D9D8>


目标：创建不同规格的FC网络
输入（张量）：输入张量
num_layers：FCNet 中的层数
h_dim（int）：隐藏单元的数量
h_fn：隐藏层的激活函数（默认值：tf.nn.relu)
o_dim (int) ：输出单元的数量
o_fn ：输出层的激活函数（默认值：无）
w_init ：权重矩阵的初始化（默认值：Xavier）
keep_prob ：保持概率 [0, 1] （如果没有，则不使用 dropout）

输入数据
dbfs ls -l dbfs:/FileStore/
文件 12845144 Losv2_311-800.data-00000-of-00001
文件 1756 Losv2_311-800.index
文件 330197 Losv2_311-800.meta
文件 260 Losv2_311_par.pkl
文件 137 检查点

标签： tensorflowvariablesmodelrestore